Disparities in Community-Based Rehabilitation Following Stroke: Findings From the COMPASS Trial

Our objective was to assess the extent of sociodemographic and geographic disparities in the use of community-based rehabilitation following stroke, an area that has been understudied.

We conducted an observational, secondary analysis of data from the COMPASS (Comprehensive Post-Acute Stroke Services) trial, linking data of 6754 survivors of stroke to insurance claims. Our primary outcome was receipt of physical or occupational therapy in the home or outpatient setting within 30 days of hospital discharge. Secondary outcomes included therapy receipt within 90 days, time to first visit, number of visits, care in the home versus outpatient setting, and therapy type (physical therapy and occupational therapy versus 1 type). Our exposures were sex, race, rurality of patient residence, and socioeconomic status (Medicaid insurance). We used generalized estimating equations to examine associations between exposures and outcomes, controlling for therapy need.

We found evidence of disparities in therapy receipt, setting, and type. Individuals on Medicaid (versus no Medicaid) were 8% (-0.08 [95% CI, -0.10 to -0.05]) less likely to receive therapy within 30 days, with the effect being slightly smaller for White patients (-0.05 [95% CI, -0.09 to -0.01]). Individuals living in rural versus metropolitan areas were 5% less likely (-0.05 [95% CI, -0.08 to -0.02]) to receive care within 30 days and 12% (-0.12 [95% CI, -0.22 to -0.03]) less likely to receive physical therapy and occupational therapy. Overall, we found minimal evidence of racial disparities.

Our findings add to the limited evidence on disparities in community-based rehabilitation after stroke.